The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or security vulnerabilities were detected across the analyzed files.
[DATA_EXPOSURE]: The skill correctly uses environment variables (e.g., BETTER_AUTH_SECRET, GOOGLE_CLIENT_ID) for sensitive configurations. It includes a setup script that generates a secure random secret using openssl rather than hardcoding credentials.
[EXTERNAL_DOWNLOADS]: All external references and dependencies are directed toward official documentation, well-known registries (NPM), and trusted organizations (Cloudflare, Hono, better-auth).
[COMMAND_EXECUTION]: The skill provides standard development commands for project setup, dependency installation, and database migrations. The included shell script performs routine file system and CLI operations expected for library configuration.
[INDIRECT_PROMPT_INJECTION]: The skill documents how to handle user data through database ORMs (Drizzle, Kysely), which inherently provide protection against common injection attacks like SQLi. It does not introduce surfaces for indirect instruction injection into the agent's logic.

better-auth